meaning 'lazy' or 'inactive', in reference to its inertness
Japan
Joined September 2025
- Tweets13.5K
- Following1.9K
- Followers24
- Likes40K
eMudhra 4 Argon retweeted
I dug more into Gemini’s speed issues. Almost all of it is server side. Their inference is awful.
It takes 11 seconds before it starts to respond, 20x longer than Fable.
eMudhra 4 Argon retweeted
Introducing RedstoneBench
Models optimize verifiable redstone builds to minimize blocks
Astra + Sol are currently SotA, but I haven't run Claude yet
If you have some spare Claude limits, there's a skill in the repo for contributing data, just give it the link and say, "Run Fable 5.2"
eMudhra 4 Argon retweeted
reserve your @muse.ai email handle while you can, paid subs!
also, since many people were asking, this will be out to everyone soon including free plans!
eMudhra 4 Argon retweeted
Google says Gemini Decision is the best calibrated decision model
Their reported number actually puts them 2nd
Still great but not #1
eMudhra 4 Argon retweeted
Gemini 4 Carbon is here. It's slow, awkward, stupid, and hallucinates all the time.
That said, they made a ton of progress and the team is working hard to make it great.
There's also some surprisingly useful capabilities hidden deep if you dig 👀
eMudhra 4 Argon retweeted
timeless
I am not a fan of Theo @theo but I tried T3 code after using claude and codex app and honestly it's the best. It solved my remote control problem on my wsl and vps...
eMudhra 4 Argon retweeted
Introducing Gemini Decision 1, our new frontier decision model in the Gemini API.
It is $0.10 per 1M input tokens with no additional cost. With support for image, audio, video, and PDF natively, 256K context window, & 1K questions.
You can try it for free in the API!
eMudhra 4 Argon retweeted
You can bring your ChatGPT subscription to the Warp terminal, now with support for business plans too!
Now, you can use your ChatGPT Business plan in tools like @NotionHQ, @v0, @ampcode, @kilocode, @warpdotdev, Hermes from @NousResearch, @openclaw, @opencode, and more.
Connect with Sign in with ChatGPT to get started.
eMudhra 4 Argon retweeted
Google's very first decision model is here, and it's on OpenRouter from day one.
Gemini Decision 1 answers yes/no, multiple choice, and score questions with a probability, for $0.10/M input and free output.
openrouter.ai/google/gemini-…
Gemini Decision 1 from @GoogleAIStudio is live on OpenRouter!
Google's first decision model. Hand it text, images, audio, video, or a PDF plus typed questions, and it answers yes/no, picks from up to 255 options, or scores, each with a probability. 262K context.
eMudhra 4 Argon retweeted
(1/10) Today, we’re sharing new research from @OpenAI: How does CoT monitorability (no-CoT capability and CoT controllability) evolve over the model training lifecycle?
We emerged with unsurprising but sobering findings: as we scale toward more intelligent systems, models are becoming harder to monitor through their CoT. 🧵
eMudhra 4 Argon retweeted
GOOGLE 🔥
gemini-decision-1 is out in Gemini API models.
• Input token limit of 262144
• Output token limit of 1
eMudhra 4 Argon retweeted
Meet Nex-N2.6, our latest multimodal agentic model supporting text, image, and video inputs.
Open-sourced by Shanghai Innovation Institute, the model has performed strong benchmark results:
⚡Compared with Nex-N2.5-Pro, the reasoning repetition rate falls from 16.0% to 1.6%, while the overall non-repetition rate rises from 91.0% to 97.8%
⚡Achieved 100% valid tool selection rate across 956 tested contexts, on par with GPT-6-Sol
⚡Scored 83.0 on OSWorld and 84.1 on Terminal-Bench 2.1, ahead of DeepSeek-V4.1-Flash at 71.7 and 82.7
Nex-N2.6 is built with a stronger foundation to navigate interfaces, use tools, and complete lengthy tasks without losing its place.
Try now 👇
⭐GitHub: github.com/sii-nex/Nex-N2.6
🤗HuggingFace: huggingface.co/nex-agi/Nex-N…
🔗OpenRouter API: openrouter.ai/nex-agi/nex-n2…
🔗MOSS API: platform.moss.cn/docs/models…
eMudhra 4 Argon retweeted
ERNIE 5.5 Flash Preview just landed! 🔥
Built on a new architecture designed for efficient inference, the latest open-weight foundation model in the ERNIE family delivers stronger performance in coding, 3D modeling, knowledge work and content creation.
It keeps decoding fast while posting competitive scores on Terminal-Bench, AutomationBench, and other benchmarks.
Here's the rundown:
- 280B total parameters, 14B active per token
- 1M-token context window
- Generation speed above 200 tokens/sec
Model weights will be released by November 25.
> And most importantly, dark mode!
愛してます
Introducing AA+
A Chrome extension that turns Artificial Analysis into an user-oriented analytics tool
It has:
- Subsidization data from Priors
- More customizability
- Radar charts to show strengths and weaknesses of different models
And most importantly, dark mode!
1
11
